Crucible/Fisheye(CF) is going to be used for code reviews ONLY even though Fisheye is a SVN Repo browser with limited capabilities compared with ToroiseSVN or RapidSVN. CF needs to parse the repository and then store the results in a SQL based database(MySQL in our case). This is an incredibly slow process. I don't need CF to parse the "actual" source code repository branches (i.e. /main, /trunk, /branches/release[s]/) because the code needs to be inspected before it is allowed on these branches. As a result, the only branch that needs to be parsed is the /branches/private/ subdirectories but this is massive with every designer taking a copy of /main, /trunk or /branches/release[s]/?? to work on. CF with its native/default svn plugin does not understand or emulate the lightweight directory structure of SVN and its incredibly efficient means of copying directories and only storing the deltas.
The actual/entire repository is 5.7G in size and after a day of parsing (and indicating 0% complete) of just the /branches/private/* subtree, the "FISHEYE_INST" directory is 1.2G so this could result in a CF database being considerably larger than the actual SVN repo.
Suggestions?
